$Header: /cvsroot/marf/marf/HISTORY,v 1.6 2005/06/06 17:35:30 mokhov Exp $

[See History section in the PDF manual for some additional info. TODO: merge]

This project has started as a course project for Pattern Recognition,
Fall 2002, at Concordia University (http://www.concordia.ca/),
Dept. of Computer Science (http://www.cs.concordia.ca), Montreal, Canada,
by four students:

  Ian Clément
  Serguei Mokhov
  Dimitrios Nicolacopoulos
  Stephen Sinclair

The project seems to have survived past that school term,
and is being maintained by us.

[XXX UPDATE:]

There has been three official "releases" of our framework and
and SpeakerIdentApp so far.

- Monday, February 10, 2003

  MARF 0.2.0 - Our first release after the project submission

    - new API in MARF - train()/recognize()
    - global Makefile build system
      - Compilation
      - Distros
    - TrainingSet improvments
    - new modules
        - BandpassFilter
        - MinkowskiDistance
        - More or less functional NeuralNet
        - "Random" feature extraction and classification
        - MahalanobisDistance, particular case for identity matrix
    - improved docs (a lot!)
    - code robustness via exceptions and 'final'

  SpeakerIdentApp, 0.2.0
    - Incorporates changes to MARF
    - New helpful options


- December 20, 2002

  MARF 0.1.2 - Project

  Includes what we have submitted (including an ISO CD image) as our final
  project deliverables, the MARF, SpeakerIdenApp along with voice samples,
  and the report.


- December 8, 2002

  MARF 0.1.0 - Demo

  In-class demo of our project to the professor, Dr. Suen, and Javad Sadri of
  at least what we've got working that day :-)

EOF